effectcraft-motion

An agent skill (Claude Code, Codex, OpenCode, and other agents that read SKILL.md) for building motion-graphics promos, explainers and logo stings in EffectCraft, with voiceover, music and SFX in sync.

The silent sample above is what ec.py demo builds and renders on your machine (1920×1080, 20 s, a made-up brand).

It ships five tested looks, ready-made layouts (statement, list, flow, icons, bars, phone, stats, end card…), fonts, and QA gates that catch empty frames, unreadable text, overlapping voice lines and off-beat SFX.

Quick start

npx skills add akdavid4real/effectcraft-motion -g          # installs into your agents' skills folders
python3 ~/.claude/skills/effectcraft-motion/scripts/ec.py doctor   # what you have, what's missing
python3 ~/.claude/skills/effectcraft-motion/scripts/ec.py demo     # build + render the sample above

No Node? Clone it instead:

git clone https://github.com/akdavid4real/effectcraft-motion ~/.claude/skills/effectcraft-motion

(For Codex, OpenCode and similar agents, use ~/.agents/skills/ in place of ~/.claude/skills/.)

Then ask your agent for a video, for example "Make a 60-second promo for example.com in EffectCraft".

What you need

Start with the minimum. Everything else only adds sound.

Minimum: silent videos and the demo

Tool	Install
EffectCraft (effectcraft-cli)	getartcraft.com/apps/effectcraft; set EFFECTCRAFT_CLI if it's not on PATH
Python 3.9+ with numpy	pip install numpy
ffmpeg + ffprobe	sudo apt install ffmpeg / brew install ffmpeg
fontconfig	sudo apt install fontconfig / brew install fontconfig (the kit installs its own fonts)

Optional: voiceover, music, SFX

Tool	Adds	Install
Node 18+ and a Kokoro venv	a free local voiceover and word timings (npx hyperframes tts / transcribe)	nodejs.org; python3 -m venv ~/.venvs/kokoro && ~/.venvs/kokoro/bin/pip install kokoro-onnx soundfile; export HYPERFRAMES_PYTHON=~/.venvs/kokoro/bin/python
media-use skill + heygen CLI	music and SFX by description	the HyperFrames skills; heygen CLI, then heygen auth login --oauth (free allowance)
edge-tts	another free voice (online)	pipx install edge-tts

ec.py doctor checks all of the above and prints the install command for anything missing.

Third-party dependencies

The voice and music steps use tools this project doesn't control: HyperFrames (npx hyperframes for TTS and transcription), its media-use skill, and the HeyGen catalog. Their commands, models or free allowances can change. If one breaks, the kit still works:

• Music: put your own track in assets/bgm/ and point "music" in timeline.json at it.
• SFX: save your own sounds as assets/sfx/<name>.mp3 and cue them as {"file": "<name>", "at": ...} in timeline.json. timeline.py mix places and mixes them.
• Voice: record or generate the lines with any tool. Word timings can come from any transcriber that outputs start and end times (audio_meta.json, shape in SKILL.md).
• No audio at all: time scenes with "in+0.5" offsets instead of spoken words, as demo/ does.

Building, QA and rendering depend only on EffectCraft, Python, numpy and ffmpeg.

Update

npx skills update effectcraft-motion        # or: git -C ~/.claude/skills/effectcraft-motion pull

Author

AK David (@akdavid4real). MIT licence; the bundled fonts are OFL (see fonts/LICENSE.md).